Professor Curtin's research is predominantly in the area of multiscale modeling of the mechanical behavior of materials. Within this theme lie a number of specific research areas:
(i) Multiscale modeling of fracture, from continuum down to atomistic and quantum scales
(ii) Multiscale modeling of fiber-reinforced composites
(iii) Atomistic modeling of deformation of nanoscale structures
(iv) Development of nanotube ceramic matrix composites including atomistic modeling of multiwall nanotubes including interwall coupling
(v) Discrete-dislocation modeling of fracture in thin metal films and fatigue crack growth emanating from small inclusions in structural metals such as Al
(vi) Strengthening and rate-dependent deformation in metals, with particular emphases on solute strengthening and on dynamic strain aging in Al-Mg and related alloys.
Many of these efforts are coupled to experimental studies carried out by collaborators at Brown and elsewhere.
A more in-depth discussion of several projects is provided below.
1. Atomistic/Continuum Coupling. This research focuses on computational methods to connect atomistic-scale and continuum-scale descriptions of material behavior within a unified framework to predict deformation and fracture. An unique feature of this work is that important phenomena are occurring at multiple scales. In the Coupled-Atomistic/Discrete-Dislocation (CADD) model developed by Curtin and co-workers, the continuum region deforms according to a discrete-dislocation plasticity description, which captures the small-scale material behavior absent in standard continuum models of plasticity.
2. Discrete/Continuum Coupling. Material behavior is often controlled by the collective motion of discrete defects, such as dislocations or impurity atoms, at scales above those that can be modeled by fully-atomistic methods. Discrete-defect models, such as the discrete-dislocation model of Needleman and coworkers, provide a framework for treating such problems. However, such approaches remain limited in the size scales that can be handled numerically, and are not feasible for direct implementation into engineering design. Our research is thus aimed at coupling such discrete models directly to "continuum models" in which there are no defects but rather simply field equations and constitutive laws relating, for instance, stress and strain - this is the standard domain of engineering design. Such a coupling is subtle since one method contains explicit discrete defects and the other does not; averaging methods and proper interfaces between regions of space described by the different methods must be developed. We have currently accomplished this for both impurity diffusion and dislocation plasticity.
3. Solute strengthening in Aluminum alloys. Al alloys are lightweight materials with a number of attractive features for many aerospace and automotive applications. The introduction of solute atoms, such as Mg, into Al provides important strengthening of the material but creates problems in forming the material at room temperature. The problems arise because the Mg (or other solute atoms) can diffuse, leading to a time-dependent strengthening referred to as Dynamic Strain Aging (DSA). The specific atomistic-scale mechanisms of strengthening and DSA phenomena are qualitatively understood but quantitative prediction has remained elusive and hence strategies for material design are not based on a solid foundation. Our research in this area has thus focussed on atomistic modelling of the interactions between Al dislocations and Mg, the most important solute addition in commercial alloys, to uncover the mechanisms of strengthening and DSA. We have recently found that DSA stems from a very local diffusion of Mg atoms within the very core of Al dislocations, which differs from the traditional view yet leads to quantitative predictions that agree with experimental data on Al-Mg alloys. Based on this atomistic insight, we have developed higher-scale constitutive models that predict the strain-rate, temperature, strain, and solute concentration, dependence of the material response. These models are being implemented in continuum-level analysis of test samples to understand the role of dynamic strain aging in driving low ductility and Portevin-LeChatelier instabilities.